Future Alli Actual Use, Label Comprehension Commitments Likely

GlaxoSmithKline could be asked to conduct more actual use research to support the possible transition of the weight loss drug Xenical (orlistat) from prescription to over-the-counter sale.

The switch proposal was reviewed by FDA's Nonprescription Drugs and Endocrinologic & Metabolic Drugs advisory committees Jan. 23. The committees recommended OTC approval for the drug in an 11-3 vote (see ¹ (Also see "Alli Makes Passing Grade On Safety; Advisory Committee Debates Benefit" - Pink Sheet, 30 Jan, 2006.)).

However, many panel members expressed concern about the strength of the actual use and label comprehension studies submitted by GSK in support of the switch application, citing inadequate power, duration and demographic diversity.

The actual use study for Alli , the proposed brand name for the 60 mg OTC version of Xenical, took place over a four-month period, although the drug is indicated for use for up to six months to aid weight loss, panel members pointed out.

According to the label, Alli can be used for up to six months, after which a doctor should be consulted. The four-month actual use study did not provide data on consumers who may exceed the six-month course.

"I think the actual use study should be longer, I would say at least 12 months, to allow for data that would give insight into the repeated use issue," NDAC member Ruth Parker, MD, Emory University, Department of Medicine, stated.

Other panel members took issue with the size of the actual use study, which included 681 eligible subjects and 237 evaluable users.

"I would really like to see a much bigger study of actual use with long-term follow-up data on persistence of weight loss," voting consultant Marie Griffin, MD, Vanderbilt University, commented.

According to GSK, the potential market for an OTC version of Xenical would include 5 mil.-6 mil. consumers.

The fact that the label used in the actual use trial had not been previously tested for comprehension generated some concerns; the language used in the label comprehension study was revised based on the results of the actual use study.

"In terms of the order of things, [the studies were] sort of done backwards. Generally, we recommend sponsors do label comprehension evaluation before an actual use study is done," Andrea Leonard-Segal, MD, Acting Director of CDER's Division of Nonprescription Clinical Evaluation, explained.

The actual use study label instructed consumers with high blood pressure, high cholesterol, more than 30 pounds to lose, or those following a doctor-recommended diet to ask a doctor before using orlistat; correct self-selection rates for users with these contraindications averaged about 41%.

GSK subsequently deleted these four warnings from the label submitted in the NDA and recalculated the self-selection rates.

This revision of the actual use trial statistically improved the self-selection results. Some panel members were uncomfortable with asserting that consumers made correct self-selection decisions based on these reevaluated results.

"We're supposed to take it on faith" that consumers will self-select correctly, NDAC member Mary Tinetti, MD, Yale University School of Medicine, said.

"We really have no comfort that they're going to select correctly," she added.

Voting consultant Sonia Patten, PhD, Macalester College, Department of Anthropology, voiced a need for a new actual use study that would employ the revised selection criteria.

Another issue cited by the panel was that the demographics of the study population did not reflect the actual U.S. population that could be exposed to the drug in an OTC setting.

In the actual use trial, 82% of the study's users were Caucasian, while Hispanics and African Americans made up 6% and 3% of the user population, respectively.

The label comprehension study was demographically similar, although no Hispanics were represented.

"I worry about the tremendous under-representation of ethnic minorities in the label comprehension and the actual use studies," Patten said.

"If we think of this from a public health or population point of view, some of those populations are at enhanced risk for obesity and all of the adverse outcomes of overweight," she stated.

The committee also discussed consumers' comprehension of and compliance with instructions about multivitamin use during orlistat therapy.

According to the labels used in both studies, consumers were directed to take a multivitamin two hours before or after taking orlistat to prevent deficiencies of fat soluble vitamins that could occur with orlistat use.

Although 75%-86% of users in the actual use study took a multivitamin, only 38%-53% of users timed the vitamin correctly, FDA stated in a review (² (Also see "Self-Selection Doubts Are No Friend To Alli As Xenical Heads For Switch Panel" - Pink Sheet, 23 Jan, 2006.), p. 5).

NDAC Chair Alastair Wood, MD, Vanderbilt University School of Medicine, felt that the vitamin issue "wasn't well dealt with."

"I think that there should be a way of making sure that multivitamins are actually delivered to the patients, preferably in the package," he stated.

GSK cited the difficulty in coordinating expiration dates of Alli and multivitamins, as well as consumers' preference to choose their own multivitamins, as impediments to a co-packaging plan.

Wood suggested these problems were surmountable.

In the case of a positive outcome for GSK, it is likely the firm will be required to conduct additional studies in an OTC environment.

The agency generally follows the recommendations of its advisory committees for drug approvals, but notably went against a strong panel recommendation supporting the switch of the emergency contraceptive Plan B .

One potential outcome would be that FDA would choose to issue an "approvable" letter for Alli pending new actual use data, given the concerns of the committees.

In another scenario, FDA may require more actual use data following approval as part of GSK's post-marketing commitments.

According to Wood, a "significant number" of Phase IV post-marketing studies must be completed in a "reasonable" period of time.

"But most importantly, the results of these studies ought to inform future changes to the labeling and use," he asserted.

- Jessica Lake